| 2 |
|
from crab_logger import Logger |
| 3 |
|
from crab_exceptions import * |
| 4 |
|
from crab_util import * |
| 5 |
+ |
import math |
| 6 |
|
import common |
| 7 |
|
import PsetManipulator |
| 8 |
|
|
| 41 |
|
|
| 42 |
|
### collect Data cards |
| 43 |
|
try: |
| 44 |
< |
# self.owner = cfg_params['CMSSW.owner'] |
| 45 |
< |
# log.debug(6, "CMSSW::CMSSW(): owner = "+self.owner) |
| 46 |
< |
# self.dataset = cfg_params['CMSSW.dataset'] |
| 47 |
< |
self.datasetPath = cfg_params['CMSSW.datasetpath'] |
| 48 |
< |
log.debug(6, "CMSSW::CMSSW(): datasetPath = "+self.datasetPath) |
| 44 |
> |
tmp = cfg_params['CMSSW.datasetpath'] |
| 45 |
> |
log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp) |
| 46 |
> |
if string.lower(tmp)=='none': |
| 47 |
> |
self.datasetPath = None |
| 48 |
> |
self.selectNoInput = 1 |
| 49 |
> |
else: |
| 50 |
> |
self.datasetPath = tmp |
| 51 |
> |
self.selectNoInput = 0 |
| 52 |
|
except KeyError: |
| 49 |
– |
# msg = "Error: owner and/or dataset not defined " |
| 53 |
|
msg = "Error: datasetpath not defined " |
| 54 |
|
raise CrabException(msg) |
| 55 |
|
|
| 56 |
|
# ML monitoring |
| 57 |
|
# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM |
| 58 |
< |
datasetpath_split = self.datasetPath.split("/") |
| 59 |
< |
self.setParam_('dataset', datasetpath_split[1]) |
| 60 |
< |
self.setParam_('owner', datasetpath_split[-1]) |
| 58 |
> |
if not self.datasetPath: |
| 59 |
> |
self.setParam_('dataset', 'None') |
| 60 |
> |
self.setParam_('owner', 'None') |
| 61 |
> |
else: |
| 62 |
> |
datasetpath_split = self.datasetPath.split("/") |
| 63 |
> |
self.setParam_('dataset', datasetpath_split[1]) |
| 64 |
> |
self.setParam_('owner', datasetpath_split[-1]) |
| 65 |
> |
|
| 66 |
|
self.setTaskid_() |
| 67 |
|
self.setParam_('taskId', self.cfg_params['taskId']) |
| 68 |
|
|
| 61 |
– |
|
| 62 |
– |
|
| 63 |
– |
|
| 69 |
|
self.dataTiers = [] |
| 65 |
– |
# try: |
| 66 |
– |
# tmpDataTiers = string.split(cfg_params['CMSSW.data_tier'],',') |
| 67 |
– |
# for tmp in tmpDataTiers: |
| 68 |
– |
# tmp=string.strip(tmp) |
| 69 |
– |
# self.dataTiers.append(tmp) |
| 70 |
– |
# pass |
| 71 |
– |
# pass |
| 72 |
– |
# except KeyError: |
| 73 |
– |
# pass |
| 74 |
– |
# log.debug(6, "Cmssw::Cmssw(): dataTiers = "+str(self.dataTiers)) |
| 70 |
|
|
| 71 |
|
## now the application |
| 72 |
|
try: |
| 112 |
|
|
| 113 |
|
# script_exe file as additional file in inputSandbox |
| 114 |
|
try: |
| 115 |
< |
self.scriptExe = cfg_params['USER.script_exe'] |
| 116 |
< |
self.additional_inbox_files.append(self.scriptExe) |
| 115 |
> |
self.scriptExe = cfg_params['USER.script_exe'] |
| 116 |
> |
self.additional_inbox_files.append(self.scriptExe) |
| 117 |
> |
if self.scriptExe != '': |
| 118 |
> |
if not os.path.isfile(self.scriptExe): |
| 119 |
> |
msg ="WARNING. file "+self.scriptExe+" not found" |
| 120 |
> |
raise CrabException(msg) |
| 121 |
|
except KeyError: |
| 122 |
|
pass |
| 124 |
– |
if self.scriptExe != '': |
| 125 |
– |
if os.path.isfile(self.scriptExe): |
| 126 |
– |
pass |
| 127 |
– |
else: |
| 128 |
– |
log.message("WARNING. file "+self.scriptExe+" not found") |
| 129 |
– |
sys.exit() |
| 123 |
|
|
| 124 |
|
## additional input files |
| 125 |
|
try: |
| 134 |
|
except KeyError: |
| 135 |
|
pass |
| 136 |
|
|
| 137 |
+ |
# files per job |
| 138 |
|
try: |
| 139 |
|
self.filesPerJob = int(cfg_params['CMSSW.files_per_jobs']) #Daniele |
| 140 |
+ |
self.selectFilesPerJob = 1 |
| 141 |
|
except KeyError: |
| 142 |
< |
self.filesPerJob = 1 |
| 142 |
> |
self.filesPerJob = 0 |
| 143 |
> |
self.selectFilesPerJob = 0 |
| 144 |
|
|
| 145 |
< |
## Max event will be total_number_of_events ??? Daniele |
| 145 |
> |
## Events per job |
| 146 |
|
try: |
| 147 |
< |
self.maxEv = cfg_params['CMSSW.event_per_job'] |
| 147 |
> |
self.eventsPerJob =int( cfg_params['CMSSW.events_per_job']) |
| 148 |
> |
self.selectEventsPerJob = 1 |
| 149 |
|
except KeyError: |
| 150 |
< |
self.maxEv = "-1" |
| 151 |
< |
## |
| 150 |
> |
self.eventsPerJob = -1 |
| 151 |
> |
self.selectEventsPerJob = 0 |
| 152 |
> |
|
| 153 |
> |
## number of jobs |
| 154 |
> |
try: |
| 155 |
> |
self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs']) |
| 156 |
> |
self.selectNumberOfJobs = 1 |
| 157 |
> |
except KeyError: |
| 158 |
> |
self.theNumberOfJobs = 0 |
| 159 |
> |
self.selectNumberOfJobs = 0 |
| 160 |
> |
|
| 161 |
> |
## source seed for pythia |
| 162 |
> |
try: |
| 163 |
> |
self.sourceSeed = int(cfg_params['CMSSW.pythia_seed']) |
| 164 |
> |
except KeyError: |
| 165 |
> |
self.sourceSeed = None |
| 166 |
> |
common.logger.debug(5,"No seed given") |
| 167 |
> |
|
| 168 |
> |
if not (self.selectFilesPerJob + self.selectEventsPerJob + self.selectNumberOfJobs == 1 ): |
| 169 |
> |
msg = 'Must define either files_per_jobs or events_per_job or number_of_jobs' |
| 170 |
> |
raise CrabException(msg) |
| 171 |
> |
|
| 172 |
|
try: |
| 173 |
|
self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events']) |
| 174 |
|
except KeyError: |
| 213 |
|
self.maxEvents=0 # max events available ( --> check the requested nb. of evts in Creator.py) |
| 214 |
|
self.DBSPaths={} # all dbs paths requested ( --> input to the site local discovery script) |
| 215 |
|
## Perform the data location and discovery (based on DBS/DLS) |
| 216 |
< |
self.DataDiscoveryAndLocation(cfg_params) |
| 216 |
> |
## SL: Don't if NONE is specified as input (pythia use case) |
| 217 |
> |
common.analisys_common_info['sites']=None |
| 218 |
> |
if self.datasetPath: |
| 219 |
> |
self.DataDiscoveryAndLocation(cfg_params) |
| 220 |
|
#DBSDLS-end |
| 221 |
|
|
| 222 |
|
self.tgzNameWithPath = self.getTarBall(self.executable) |
| 223 |
+ |
|
| 224 |
+ |
## Select Splitting |
| 225 |
+ |
if self.selectNoInput: self.jobSplittingNoInput() |
| 226 |
+ |
elif self.selectFilesPerJob or self.selectEventsPerJob or self.selectNumberOfJobs: self.jobSplittingPerFiles() |
| 227 |
+ |
else: |
| 228 |
+ |
msg = 'Don\'t know how to split...' |
| 229 |
+ |
raise CrabException(msg) |
| 230 |
|
|
| 231 |
< |
self.jobSplitting() #Daniele job Splitting |
| 232 |
< |
self.PsetEdit.maxEvent(self.maxEv) #Daniele |
| 233 |
< |
self.PsetEdit.inputModule("INPUT") #Daniele |
| 234 |
< |
self.PsetEdit.psetWriter(self.configFilename()) |
| 235 |
< |
|
| 236 |
< |
|
| 231 |
> |
# modify Pset |
| 232 |
> |
try: |
| 233 |
> |
if (self.datasetPath): # standard job |
| 234 |
> |
#self.PsetEdit.maxEvent(self.eventsPerJob) |
| 235 |
> |
# always process all events in a file |
| 236 |
> |
self.PsetEdit.maxEvent("-1") |
| 237 |
> |
self.PsetEdit.inputModule("INPUT") |
| 238 |
> |
|
| 239 |
> |
else: # pythia like job |
| 240 |
> |
self.PsetEdit.maxEvent(self.eventsPerJob) |
| 241 |
> |
if (self.sourceSeed) : |
| 242 |
> |
self.PsetEdit.pythiaSeed("INPUT","INPUTVTX") |
| 243 |
> |
self.PsetEdit.psetWriter(self.configFilename()) |
| 244 |
> |
except: |
| 245 |
> |
msg='Error while manipuliating ParameterSet: exiting...' |
| 246 |
> |
raise CrabException(msg) |
| 247 |
|
|
| 248 |
|
def DataDiscoveryAndLocation(self, cfg_params): |
| 249 |
|
|
| 250 |
|
common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()") |
| 251 |
|
|
| 215 |
– |
#datasetPath = "/"+self.owner+"/"+self.dataTiers[0]+"/"+self.dataset |
| 216 |
– |
|
| 252 |
|
datasetPath=self.datasetPath |
| 253 |
|
|
| 254 |
|
## TODO |
| 276 |
|
common.logger.message("Required data are :"+self.datasetPath) |
| 277 |
|
|
| 278 |
|
filesbyblock=self.pubdata.getFiles() |
| 279 |
+ |
# print filesbyblock |
| 280 |
|
self.AllInputFiles=filesbyblock.values() |
| 281 |
|
self.files = self.AllInputFiles |
| 282 |
|
|
| 247 |
– |
## TEMP |
| 248 |
– |
# self.filesTmp = filesbyblock.values() |
| 249 |
– |
# self.files = [] |
| 250 |
– |
# locPath='rfio:cmsbose2.bo.infn.it:/flatfiles/SE00/cms/fanfani/ProdTest/' |
| 251 |
– |
# locPath='' |
| 252 |
– |
# tmp = [] |
| 253 |
– |
# for file in self.filesTmp[0]: |
| 254 |
– |
# tmp.append(locPath+file) |
| 255 |
– |
# self.files.append(tmp) |
| 256 |
– |
## END TEMP |
| 257 |
– |
|
| 283 |
|
## get max number of events |
| 284 |
|
#common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents()) |
| 285 |
|
self.maxEvents=self.pubdata.getMaxEvents() ## self.maxEvents used in Creator.py |
| 304 |
|
msg = 'No sites hosting all the needed data! Exiting... ' |
| 305 |
|
raise CrabException(msg) |
| 306 |
|
|
| 307 |
< |
common.logger.message("List of Sites hosting the data : "+str(sites)) |
| 307 |
> |
common.logger.message("List of Sites ("+str(len(sites))+") hosting the data : "+str(sites)) |
| 308 |
|
common.logger.debug(6, "List of Sites: "+str(sites)) |
| 309 |
|
common.analisys_common_info['sites']=sites ## used in SchedulerEdg.py in createSchScript |
| 310 |
|
self.setParam_('TargetCE', ','.join(sites)) |
| 311 |
|
return |
| 312 |
|
|
| 313 |
< |
def jobSplitting(self): |
| 313 |
> |
def jobSplittingPerFiles(self): |
| 314 |
|
""" |
| 315 |
< |
first implemntation for job splitting |
| 316 |
< |
""" |
| 317 |
< |
# print 'eventi totali '+str(self.maxEvents) |
| 293 |
< |
# print 'eventi totali richiesti dallo user '+str(self.total_number_of_events) |
| 294 |
< |
#print 'files per job '+str(self.filesPerJob) |
| 295 |
< |
common.logger.message('Required '+str(self.filesPerJob)+' files per job ') |
| 315 |
> |
Perform job splitting based on number of files to be accessed per job |
| 316 |
> |
""" |
| 317 |
> |
common.logger.debug(5,'Splitting per input files') |
| 318 |
|
common.logger.message('Required '+str(self.total_number_of_events)+' events in total ') |
| 319 |
+ |
common.logger.message('Available '+str(self.maxEvents)+' events in total ') |
| 320 |
+ |
common.logger.message('Required '+str(self.filesPerJob)+' files per job ') |
| 321 |
+ |
common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ') |
| 322 |
+ |
common.logger.message('Required '+str(self.eventsPerJob)+' events per job') |
| 323 |
+ |
|
| 324 |
+ |
## if asked to process all events, do it |
| 325 |
+ |
if self.total_number_of_events == -1: |
| 326 |
+ |
self.total_number_of_events=self.maxEvents |
| 327 |
+ |
else: |
| 328 |
+ |
if self.total_number_of_events>self.maxEvents: |
| 329 |
+ |
common.logger.message("Asked "+str(self.total_number_of_events)+" but only "+str(self.maxEvents)+" available.") |
| 330 |
+ |
self.total_number_of_events=self.maxEvents |
| 331 |
+ |
pass |
| 332 |
|
|
| 333 |
|
## TODO: SL need to have (from DBS) a detailed list of how many events per each file |
| 334 |
|
n_tot_files = (len(self.files[0])) |
| 335 |
|
## SL: this is wrong if the files have different number of events |
| 336 |
|
evPerFile = int(self.maxEvents)/n_tot_files |
| 337 |
< |
|
| 337 |
> |
|
| 338 |
|
common.logger.debug(5,'Events per File '+str(evPerFile)) |
| 339 |
|
|
| 340 |
< |
## if asked to process all events, do it |
| 341 |
< |
if self.total_number_of_events == -1: |
| 342 |
< |
self.total_number_of_events=self.maxEvents |
| 343 |
< |
self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob) |
| 344 |
< |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for all available events '+str(self.total_number_of_events)+' events') |
| 340 |
> |
## compute job splitting parameters: filesPerJob, eventsPerJob and theNumberOfJobs |
| 341 |
> |
if self.selectFilesPerJob: |
| 342 |
> |
## user define files per event. |
| 343 |
> |
filesPerJob = self.filesPerJob |
| 344 |
> |
eventsPerJob = filesPerJob*evPerFile |
| 345 |
> |
theNumberOfJobs = int(self.total_number_of_events*1./eventsPerJob) |
| 346 |
> |
check = int(self.total_number_of_events) - (theNumberOfJobs*eventsPerJob) |
| 347 |
> |
if check > 0: |
| 348 |
> |
theNumberOfJobs +=1 |
| 349 |
> |
filesLastJob = int(check*1./evPerFile+0.5) |
| 350 |
> |
common.logger.message('Warning: last job will be created with '+str(check)+' files') |
| 351 |
> |
else: |
| 352 |
> |
filesLastJob = filesPerJob |
| 353 |
> |
|
| 354 |
> |
elif self.selectNumberOfJobs: |
| 355 |
> |
## User select the number of jobs: last might be bigger to match request of events |
| 356 |
> |
theNumberOfJobs = self.theNumberOfJobs |
| 357 |
> |
|
| 358 |
> |
eventsPerJob = self.total_number_of_events/theNumberOfJobs |
| 359 |
> |
filesPerJob = int(eventsPerJob/evPerFile) |
| 360 |
> |
if (filesPerJob==0) : filesPerJob=1 |
| 361 |
> |
check = int(self.total_number_of_events) - (int(theNumberOfJobs)*filesPerJob*evPerFile) |
| 362 |
> |
if not check == 0: |
| 363 |
> |
if check<0: |
| 364 |
> |
missingFiles = int(check/evPerFile) |
| 365 |
> |
additionalJobs = int(missingFiles/filesPerJob) |
| 366 |
> |
#print missingFiles, additionalJobs |
| 367 |
> |
theNumberOfJobs+=additionalJobs |
| 368 |
> |
common.logger.message('Warning: will create only '+str(theNumberOfJobs)+' jobs') |
| 369 |
> |
check = int(self.total_number_of_events) - (int(theNumberOfJobs)*filesPerJob*evPerFile) |
| 370 |
> |
|
| 371 |
> |
if check >0 : |
| 372 |
> |
filesLastJob = filesPerJob+int(check*1./evPerFile+0.5) |
| 373 |
> |
common.logger.message('Warning: last job will be created with '+str(filesLastJob*evPerFile)+' events') |
| 374 |
> |
else: |
| 375 |
> |
filesLastJob = filesPerJob |
| 376 |
> |
else: |
| 377 |
> |
filesLastJob = filesPerJob |
| 378 |
> |
elif self.selectEventsPerJob: |
| 379 |
> |
# SL case if asked events per job |
| 380 |
> |
## estimate the number of files per job to match the user requirement |
| 381 |
> |
filesPerJob = int(float(self.eventsPerJob)/float(evPerFile)) |
| 382 |
> |
if filesPerJob==0: filesPerJob=1 |
| 383 |
> |
common.logger.debug(5,"filesPerJob "+str(filesPerJob)) |
| 384 |
> |
if (filesPerJob==0): filesPerJob=1 |
| 385 |
> |
eventsPerJob=filesPerJob*evPerFile |
| 386 |
> |
theNumberOfJobs = int(self.total_number_of_events)/int(eventsPerJob) |
| 387 |
> |
check = int(self.total_number_of_events) - (int(theNumberOfJobs)*eventsPerJob) |
| 388 |
> |
if not check == 0: |
| 389 |
> |
missingFiles = int(check/evPerFile) |
| 390 |
> |
additionalJobs = int(missingFiles/filesPerJob) |
| 391 |
> |
if ( additionalJobs>0) : theNumberOfJobs+=additionalJobs |
| 392 |
> |
check = int(self.total_number_of_events) - (int(theNumberOfJobs)*eventsPerJob) |
| 393 |
> |
if not check == 0: |
| 394 |
> |
if (check <0 ): |
| 395 |
> |
filesLastJob = filesPerJob+int(check*1./evPerFile-0.5) |
| 396 |
> |
else: |
| 397 |
> |
theNumberOfJobs+=1 |
| 398 |
> |
filesLastJob = int(check*1./evPerFile+0.5) |
| 399 |
> |
|
| 400 |
> |
common.logger.message('Warning: last job will be created with '+str(filesLastJob*evPerFile)+' events') |
| 401 |
> |
else: |
| 402 |
> |
filesLastJob = filesPerJob |
| 403 |
> |
else: |
| 404 |
> |
filesLastJob = filesPerJob |
| 405 |
|
|
| 406 |
< |
else: |
| 407 |
< |
self.total_number_of_files = int(self.total_number_of_events/evPerFile) |
| 408 |
< |
## SL: if ask for less event than what is computed to be available on a |
| 409 |
< |
## file, process the first file anyhow. |
| 410 |
< |
if self.total_number_of_files == 0: |
| 411 |
< |
self.total_number_of_files = self.total_number_of_files + 1 |
| 406 |
> |
self.total_number_of_jobs = theNumberOfJobs |
| 407 |
> |
|
| 408 |
> |
totalEventsToBeUsed=theNumberOfJobs*filesPerJob*evPerFile |
| 409 |
> |
if not check == 0: |
| 410 |
> |
# print (theNumberOfJobs-1)*filesPerJob*evPerFile,filesLastJob*evPerFile |
| 411 |
> |
totalEventsToBeUsed=(theNumberOfJobs-1)*filesPerJob*evPerFile+filesLastJob*evPerFile |
| 412 |
|
|
| 413 |
< |
common.logger.debug(5,'N files '+str(self.total_number_of_files)) |
| 413 |
> |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created, each for '+str(filesPerJob*evPerFile)+' events, for a total of '+str(totalEventsToBeUsed)+' events') |
| 414 |
|
|
| 415 |
< |
check = 0 |
| 415 |
> |
totalFilesToBeUsed=filesPerJob*(theNumberOfJobs-1)+filesLastJob |
| 416 |
> |
|
| 417 |
> |
## set job arguments (files) |
| 418 |
> |
list_of_lists = [] |
| 419 |
> |
lastFile=0 |
| 420 |
> |
for i in range(0, int(totalFilesToBeUsed), filesPerJob)[:-1]: |
| 421 |
> |
parString = "\\{" |
| 422 |
> |
|
| 423 |
> |
lastFile=i+filesPerJob |
| 424 |
> |
params = self.files[0][i: lastFile] |
| 425 |
> |
for i in range(len(params) - 1): |
| 426 |
> |
parString += '\\\"' + params[i] + '\\\"\,' |
| 427 |
|
|
| 428 |
< |
## Compute the number of jobs |
| 429 |
< |
#self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob) |
| 430 |
< |
self.total_number_of_jobs = int(self.total_number_of_files/self.filesPerJob) |
| 325 |
< |
common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs)) |
| 428 |
> |
parString += '\\\"' + params[len(params) - 1] + '\\\"\\}' |
| 429 |
> |
list_of_lists.append([parString]) |
| 430 |
> |
pass |
| 431 |
|
|
| 432 |
< |
## is there any remainder? |
| 433 |
< |
check = int(self.total_number_of_files) - (int(self.total_number_of_jobs)*self.filesPerJob) |
| 432 |
> |
## last job |
| 433 |
> |
parString = "\\{" |
| 434 |
> |
|
| 435 |
> |
params = self.files[0][lastFile: lastFile+filesLastJob] |
| 436 |
> |
for i in range(len(params) - 1): |
| 437 |
> |
parString += '\\\"' + params[i] + '\\\"\,' |
| 438 |
> |
|
| 439 |
> |
parString += '\\\"' + params[len(params) - 1] + '\\\"\\}' |
| 440 |
> |
list_of_lists.append([parString]) |
| 441 |
> |
pass |
| 442 |
|
|
| 443 |
< |
common.logger.debug(5,'Check '+str(check)) |
| 443 |
> |
self.list_of_args = list_of_lists |
| 444 |
> |
# print self.list_of_args[0] |
| 445 |
> |
return |
| 446 |
|
|
| 447 |
< |
if check > 0: |
| 448 |
< |
self.total_number_of_jobs = self.total_number_of_jobs + 1 |
| 449 |
< |
common.logger.message('Warning: last job will be created with '+str(check)+' files') |
| 447 |
> |
def jobSplittingNoInput(self): |
| 448 |
> |
""" |
| 449 |
> |
Perform job splitting based on number of event per job |
| 450 |
> |
""" |
| 451 |
> |
common.logger.debug(5,'Splitting per events') |
| 452 |
> |
common.logger.message('Required '+str(self.eventsPerJob)+' events per job ') |
| 453 |
> |
common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ') |
| 454 |
> |
common.logger.message('Required '+str(self.total_number_of_events)+' events in total ') |
| 455 |
|
|
| 456 |
< |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for a total of '+str((self.total_number_of_jobs-1)*self.filesPerJob*evPerFile + check*evPerFile)+' events') |
| 457 |
< |
pass |
| 456 |
> |
if (self.total_number_of_events < 0): |
| 457 |
> |
msg='Cannot split jobs per Events with "-1" as total number of events' |
| 458 |
> |
raise CrabException(msg) |
| 459 |
> |
|
| 460 |
> |
if (self.selectEventsPerJob): |
| 461 |
> |
self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob) |
| 462 |
> |
elif (self.selectNumberOfJobs) : |
| 463 |
> |
self.total_number_of_jobs = self.theNumberOfJobs |
| 464 |
> |
self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs) |
| 465 |
> |
|
| 466 |
> |
common.logger.debug(5,'N jobs '+str(self.total_number_of_jobs)) |
| 467 |
> |
|
| 468 |
> |
# is there any remainder? |
| 469 |
> |
check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob) |
| 470 |
> |
|
| 471 |
> |
common.logger.debug(5,'Check '+str(check)) |
| 472 |
> |
|
| 473 |
> |
common.logger.message(str(self.total_number_of_jobs)+' jobs will be created, each for '+str(self.eventsPerJob)+' for a total of '+str(self.total_number_of_jobs*self.eventsPerJob)+' events') |
| 474 |
> |
if check > 0: |
| 475 |
> |
common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but will do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob)) |
| 476 |
|
|
| 339 |
– |
list_of_lists = [] |
| 340 |
– |
for i in xrange(0, int(n_tot_files), self.filesPerJob): |
| 341 |
– |
list_of_lists.append(self.files[0][i: i+self.filesPerJob]) |
| 477 |
|
|
| 478 |
< |
self.list_of_files = list_of_lists |
| 479 |
< |
|
| 478 |
> |
# argument is seed number.$i |
| 479 |
> |
self.list_of_args = [] |
| 480 |
> |
for i in range(self.total_number_of_jobs): |
| 481 |
> |
if (self.sourceSeed): |
| 482 |
> |
self.list_of_args.append([(str(self.sourceSeed)+str(i))]) |
| 483 |
> |
else: |
| 484 |
> |
self.list_of_args.append([str(i)]) |
| 485 |
> |
#print self.list_of_args |
| 486 |
> |
|
| 487 |
|
return |
| 488 |
|
|
| 489 |
|
def split(self, jobParams): |
| 491 |
|
common.jobDB.load() |
| 492 |
|
#### Fabio |
| 493 |
|
njobs = self.total_number_of_jobs |
| 494 |
< |
filelist = self.list_of_files |
| 494 |
> |
arglist = self.list_of_args |
| 495 |
|
# create the empty structure |
| 496 |
|
for i in range(njobs): |
| 497 |
|
jobParams.append("") |
| 498 |
|
|
| 499 |
|
for job in range(njobs): |
| 500 |
< |
jobParams[job] = filelist[job] |
| 500 |
> |
jobParams[job] = arglist[job] |
| 501 |
> |
# print str(arglist[job]) |
| 502 |
> |
# print jobParams[job] |
| 503 |
|
common.jobDB.setArguments(job, jobParams[job]) |
| 504 |
|
|
| 505 |
|
common.jobDB.save() |
| 506 |
|
return |
| 507 |
|
|
| 508 |
|
def getJobTypeArguments(self, nj, sched): |
| 509 |
< |
params = common.jobDB.arguments(nj) |
| 510 |
< |
#print params |
| 511 |
< |
parString = "\\{" |
| 512 |
< |
|
| 369 |
< |
for i in range(len(params) - 1): |
| 370 |
< |
parString += '\\\"' + params[i] + '\\\"\,' |
| 371 |
< |
|
| 372 |
< |
parString += '\\\"' + params[len(params) - 1] + '\\\"\\}' |
| 373 |
< |
return parString |
| 509 |
> |
result = '' |
| 510 |
> |
for i in common.jobDB.arguments(nj): |
| 511 |
> |
result=result+str(i)+" " |
| 512 |
> |
return result |
| 513 |
|
|
| 514 |
|
def numberOfJobs(self): |
| 515 |
|
# Fabio |
| 377 |
– |
|
| 516 |
|
return self.total_number_of_jobs |
| 379 |
– |
|
| 380 |
– |
|
| 517 |
|
|
| 518 |
|
def checkBlackList(self, allSites): |
| 519 |
|
if len(self.reCEBlackList)==0: return allSites |
| 653 |
|
txt += ' echo "JOB_EXIT_STATUS = 10016"\n' |
| 654 |
|
txt += ' echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n' |
| 655 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 656 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 657 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 658 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 659 |
|
txt += ' exit 1\n' |
| 660 |
|
txt += ' fi\n' |
| 661 |
|
txt += '\n' |
| 675 |
|
txt += ' echo "JOB_EXIT_STATUS = 10034"\n' |
| 676 |
|
txt += ' echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n' |
| 677 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 678 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 679 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 680 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 681 |
|
## OLI_Daniele |
| 682 |
|
txt += ' if [ $middleware == OSG ]; then \n' |
| 683 |
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 688 |
|
txt += ' echo "JOB_EXIT_STATUS = 10018"\n' |
| 689 |
|
txt += ' echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n' |
| 690 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 691 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 692 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 693 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 694 |
|
txt += ' fi\n' |
| 695 |
|
txt += ' fi \n' |
| 696 |
|
txt += ' exit 1 \n' |
| 711 |
|
txt += ' echo "JOB_EXIT_STATUS = 50113"\n' |
| 712 |
|
txt += ' echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n' |
| 713 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 714 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 715 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 716 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 717 |
|
## OLI_Daniele |
| 718 |
|
txt += ' if [ $middleware == OSG ]; then \n' |
| 719 |
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 724 |
|
txt += ' echo "JOB_EXIT_STATUS = 50114"\n' |
| 725 |
|
txt += ' echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n' |
| 726 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 727 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 728 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 729 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 730 |
|
txt += ' fi\n' |
| 731 |
|
txt += ' fi \n' |
| 732 |
|
txt += " exit 1\n" |
| 737 |
|
job = common.job_list[nj] |
| 738 |
|
pset = os.path.basename(job.configFilename()) |
| 739 |
|
txt += '\n' |
| 740 |
< |
txt += 'InputFiles=$2\n' |
| 741 |
< |
txt += 'echo "<$InputFiles>"\n' |
| 742 |
< |
#txt += 'echo sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' \n' |
| 743 |
< |
txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n' |
| 744 |
< |
#txt += 'sed "s#{\'INPUT\'}#${InputFiles}#" $RUNTIME_AREA/'+pset+' > pset1.cfg\n' |
| 740 |
> |
if (self.datasetPath): # standard job |
| 741 |
> |
txt += 'InputFiles=$2\n' |
| 742 |
> |
txt += 'echo "Inputfiles:<$InputFiles>"\n' |
| 743 |
> |
txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n' |
| 744 |
> |
else: # pythia like job |
| 745 |
> |
if (self.sourceSeed): |
| 746 |
> |
txt += 'Seed=$2\n' |
| 747 |
> |
txt += 'echo "Seed: <$Seed>"\n' |
| 748 |
> |
txt += 'sed "s#INPUT#$Seed#" $RUNTIME_AREA/'+pset+' > pset.cfg\n' |
| 749 |
> |
else: |
| 750 |
> |
txt += '# Copy untouched pset\n' |
| 751 |
> |
txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n' |
| 752 |
> |
|
| 753 |
|
|
| 754 |
|
if len(self.additional_inbox_files) > 0: |
| 755 |
|
for file in self.additional_inbox_files: |
| 792 |
|
txt += ' cd $RUNTIME_AREA\n' |
| 793 |
|
txt += ' /bin/rm -rf $WORKING_DIR\n' |
| 794 |
|
txt += ' if [ -d $WORKING_DIR ] ;then\n' |
| 795 |
< |
txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n' |
| 796 |
< |
txt += ' echo "JOB_EXIT_STATUS = 50999"\n' |
| 797 |
< |
txt += ' echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n' |
| 798 |
< |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 795 |
> |
txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n' |
| 796 |
> |
txt += ' echo "JOB_EXIT_STATUS = 50999"\n' |
| 797 |
> |
txt += ' echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n' |
| 798 |
> |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 799 |
> |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 800 |
> |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 801 |
> |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 802 |
|
txt += ' fi\n' |
| 803 |
|
txt += ' fi \n' |
| 804 |
|
txt += ' \n' |
| 876 |
|
txt += '\n' |
| 877 |
|
txt += '# check output file\n' |
| 878 |
|
txt += 'ls '+fileWithSuffix+'\n' |
| 879 |
< |
txt += 'exe_result=$?\n' |
| 880 |
< |
txt += 'if [ $exe_result -ne 0 ] ; then\n' |
| 881 |
< |
txt += ' echo "ERROR: No output file to manage"\n' |
| 882 |
< |
txt += ' echo "JOB_EXIT_STATUS = $exe_result"\n' |
| 883 |
< |
txt += ' echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n' |
| 884 |
< |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 879 |
> |
txt += 'ls_result=$?\n' |
| 880 |
> |
#txt += 'exe_result=$?\n' |
| 881 |
> |
txt += 'if [ $ls_result -ne 0 ] ; then\n' |
| 882 |
> |
txt += ' echo "ERROR: Problem with output file"\n' |
| 883 |
> |
#txt += ' echo "JOB_EXIT_STATUS = $exe_result"\n' |
| 884 |
> |
#txt += ' echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n' |
| 885 |
> |
#txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 886 |
|
### OLI_DANIELE |
| 887 |
|
if common.scheduler.boss_scheduler_name == 'condor_g': |
| 888 |
|
txt += ' if [ $middleware == OSG ]; then \n' |
| 896 |
|
txt += 'cd $RUNTIME_AREA\n' |
| 897 |
|
file_list=file_list[:-1] |
| 898 |
|
txt += 'file_list="'+file_list+'"\n' |
| 899 |
+ |
txt += 'cd $RUNTIME_AREA\n' |
| 900 |
|
### OLI_DANIELE |
| 901 |
|
txt += 'if [ $middleware == OSG ]; then\n' |
| 902 |
|
txt += ' cd $RUNTIME_AREA\n' |
| 907 |
|
txt += ' echo "JOB_EXIT_STATUS = 60999"\n' |
| 908 |
|
txt += ' echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n' |
| 909 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 910 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 911 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 912 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 913 |
|
txt += ' fi\n' |
| 914 |
|
txt += 'fi\n' |
| 915 |
|
txt += '\n' |
| 940 |
|
return job requirements to add to jdl files |
| 941 |
|
""" |
| 942 |
|
req = '' |
| 943 |
+ |
if common.analisys_common_info['sw_version']: |
| 944 |
+ |
req='Member("VO-cms-' + \ |
| 945 |
+ |
common.analisys_common_info['sw_version'] + \ |
| 946 |
+ |
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)' |
| 947 |
|
if common.analisys_common_info['sites']: |
| 777 |
– |
if common.analisys_common_info['sw_version']: |
| 778 |
– |
req='Member("VO-cms-' + \ |
| 779 |
– |
common.analisys_common_info['sw_version'] + \ |
| 780 |
– |
'", other.GlueHostApplicationSoftwareRunTimeEnvironment)' |
| 948 |
|
if len(common.analisys_common_info['sites'])>0: |
| 949 |
|
req = req + ' && (' |
| 950 |
|
for i in range(len(common.analisys_common_info['sites'])): |
| 979 |
|
txt += ' echo "JOB_EXIT_STATUS = 10020"\n' |
| 980 |
|
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n' |
| 981 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 982 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 983 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 984 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 985 |
|
txt += ' exit 1\n' |
| 986 |
|
txt += '\n' |
| 987 |
|
txt += ' echo "Remove working directory: $WORKING_DIR"\n' |
| 992 |
|
txt += ' echo "JOB_EXIT_STATUS = 10017"\n' |
| 993 |
|
txt += ' echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n' |
| 994 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 995 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 996 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 997 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 998 |
|
txt += ' fi\n' |
| 999 |
|
txt += '\n' |
| 1000 |
|
txt += ' exit 1\n' |
| 1018 |
|
txt += ' echo "JOB_EXIT_STATUS = 10031" \n' |
| 1019 |
|
txt += ' echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n' |
| 1020 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1021 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1022 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1023 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1024 |
|
txt += ' exit 1\n' |
| 1025 |
|
txt += ' else\n' |
| 1026 |
|
txt += ' echo "Sourcing environment... "\n' |
| 1029 |
|
txt += ' echo "JOB_EXIT_STATUS = 10020"\n' |
| 1030 |
|
txt += ' echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n' |
| 1031 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1032 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1033 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1034 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1035 |
|
txt += ' exit 1\n' |
| 1036 |
|
txt += ' fi\n' |
| 1037 |
|
txt += ' echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n' |
| 1042 |
|
txt += ' echo "JOB_EXIT_STATUS = 10032"\n' |
| 1043 |
|
txt += ' echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n' |
| 1044 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1045 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1046 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1047 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1048 |
|
txt += ' exit 1\n' |
| 1049 |
|
txt += ' fi\n' |
| 1050 |
|
txt += ' fi\n' |
| 1061 |
|
txt += ' echo "JOB_EXIT_STATUS = 10033"\n' |
| 1062 |
|
txt += ' echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n' |
| 1063 |
|
txt += ' dumpStatus $RUNTIME_AREA/$repo\n' |
| 1064 |
+ |
txt += ' rm -f $RUNTIME_AREA/$repo \n' |
| 1065 |
+ |
txt += ' echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n' |
| 1066 |
+ |
txt += ' echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n' |
| 1067 |
|
txt += ' exit 1\n' |
| 1068 |
|
txt += ' fi\n' |
| 1069 |
|
txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n' |
